The aim of this experiment is to investigate whether the mechanisms by which glucocorticoids and food intake stimulate leptin in vivo involve change in the leptin diurnal cycle, and whether this effect is different between lean and obese human subjects. To answer this question, the effect of dexamethasone (0 and 2 mg) on 24-hour serum leptin will be studied in lean and obese human subjects under fed and fasting conditions. We hypothesize that: 1) dexamethasone alone, under fasting condition, will increase serum leptin levels at night and induce the nocturnal peak of leptin; 2) the cumulative effect of dexamethasone and food will increase the amplitude of the nocturnal peak of leptin, and interfere with the timing of this peak.